Enterprise Distributed Application Service (EDAS) integrates with Application Real-Time Monitoring Service (ARMS). Track application health, pinpoint slow or failing interfaces, capture call parameters for debugging, and set up alerts -- all from the EDAS console.
Quick navigation
| You want to... | Go to | Details |
|---|---|---|
| Check overall application health | Application Overview | View application overview |
| Monitor pod-level CPU, memory, and network | Prometheus | View Prometheus monitoring metrics |
| Inspect JVM heap, GC, and thread behavior | Instance Details > JVM monitoring | JVM monitoring |
| Capture a memory snapshot for analysis | Instance Details > Memory snapshots | Memory snapshots |
| Track host-level CPU, disk, load, and network | Instance Details > Host monitoring | Host monitoring |
| Analyze API call details, SQL, and dependencies | Service Details | Service and API monitoring |
| Diagnose a live performance issue | Application Diagnosis > Real-time diagnostics | Real-time diagnostics |
| Investigate application exceptions | Application Diagnosis > Exception analysis | Exception analysis |
| Profile thread CPU usage and identify thread issues | Application Diagnosis > Thread profiling | Thread profiling |
| Get notified when metrics cross a threshold | Alerts | Set up alerts |
Monitoring capabilities
EDAS monitors your application at multiple levels -- from overall health down to individual JVM threads.
Application overview
The Application Overview page displays the health of your application and its Kubernetes cluster:
Cluster information: region, microservices namespace, and cluster status
Overall metrics: total request count and average response time
Service metrics: performance of provided services and dependent services
System metrics: CPU utilization and memory usage
Diagnosis report: application diagnosis report
For details, see View application overview.
Prometheus dashboards
Preset Prometheus dashboards display pod-level metrics including basic information, CPU, memory, and network. Adjust the time interval and refresh rate to focus on a specific period.
For details, see View Prometheus monitoring metrics.
Instance-level monitoring
Drill into individual instances to inspect JVM internals, capture memory snapshots, and monitor host resources.
| Feature | What it tracks | Details |
|---|---|---|
| JVM monitoring | Heap memory, non-heap memory, direct buffer, memory-mapped buffer, garbage collections (GCs), and JVM threads | JVM monitoring |
| Memory snapshots | Detailed memory usage beyond what JVM charts show. Create a snapshot task to capture specifics and review them in logs. | Memory snapshots |
| Host monitoring | CPU, memory, disk, load, network traffic, and network packets | Host monitoring |
Service and API monitoring
Track the details of every API call your application handles:
SQL analysis and NoSQL analysis
Exception analysis and error analysis
Upstream and downstream service dependencies
Interface snapshots
For details, see Service and API monitoring.
Application diagnosis
Use diagnosis tools to identify performance issues or investigate failures:
| Tool | When to use | Details |
|---|---|---|
| Real-time diagnostics | Monitor application performance over a short period and pinpoint root causes as issues happen | Real-time diagnostics |
| Exception analysis | Review exception counts by type and by port to identify patterns | Exception analysis |
| Thread profiling | Analyze per-thread CPU consumption and thread type distribution. Method stacks are recorded and aggregated every 5 minutes, so you can trace code execution paths and locate thread issues. | Thread profiling |
Set up alerts
Define alert rules for specific monitored objects. When a rule triggers, EDAS sends notifications to the contact group you specify through your chosen channel.
To configure alerting:
Create contacts -- Add the people who should receive alert notifications. See Create a contact.
Create a contact group -- Group contacts so you can target notifications efficiently. See Create a contact group.
Create alert rules -- Define the conditions that trigger notifications. See Create an application monitoring alert rule.
After alerts are active, manage them directly in a DingTalk group. See Manage alerts in a DingTalk group.
For a complete walkthrough, see Create application monitoring alert rules and process alert notifications.